英文摘要 |
Image analysis technology is employed in this work for concrete surface strain field measurement of a post-tensioned RC-wall cyclic test. The surface strain fields of each cyclic control step are estimated by analyzing a pair of images taken by two digital cameras. Camera calibration is carried out using a user friendly MATLAB toolbox called Bouguet’s toolbox. Image distortion effect is considered using cameras’ intrinsic parameters. Stereo 3D image analysis is performed to render the 3D geometry of the measurement concrete surface. Digital Image Correlation(DIC) using template matching technique is adopted to trace the displacement change of grids on the measurement surface using a widely known high performance open source library for computer vision called OpenCV. The strain fields are estimated by the traced displacement fields. The measurement error is preliminarily estimated. This report is a four-page short version of a progress report of the first year work on image-based strain measurement [13]. A similar publication of this work has been published at The Eleventh East Asia-Pacific Conference on Structural Engineering & Construction(EASEC-11), which was held on November 19-21, 2008, Taipei, Taiwan [12]. Further details of the upcoming work will be written in the near future. |